SAA-C03 Practice Question 35: ASG Multi AZ Distribution
AWS Certified Solutions Architect – Associate (SAA-C03) · Test 4 · question 5 of 10
A web tier runs six EC2 instances in an Auto Scaling group behind an Application Load Balancer. The group is configured with a single subnet in one Availability Zone. During a recent AZ impairment the entire tier became unavailable. The business requires the application to survive the loss of one Availability Zone with no reduction in serving capacity, and to do so automatically.
What should the solutions architect change?
Show answer and explanation
Answer:
An Auto Scaling group spanning several Availability Zones balances its instances across them and launches replacements in the healthy zones when one becomes impaired. Spreading across three AZs and sizing the group so that the instances in any one zone are spare capacity means losing a zone costs a third of the fleet while the remaining two still serve the full load. The ALB, which also needs subnets in those zones, stops routing to targets that fail health checks. All of this happens without human intervention.
Option A adds capacity inside the single point of failure. Twelve instances in one Availability Zone still go away together when that zone is impaired.
Option B can work but is a manual-feeling design the service already provides. Two independent groups means two things to scale and deploy to, and Route 53 failover adds DNS TTL delay, where a multi-AZ group with an ALB shifts traffic in seconds.
Option C misunderstands both features. Termination protection prevents API-initiated termination and does not keep an instance alive through an AZ failure, and Elastic IPs do not make an instance resilient.
Reference: Auto Scaling benefits and Availability Zone distribution